Walter's Blog

Author: waltergavarrete26

  • Copilot Studio

    What if you could build an AI-powered assistant for your company — one that answers questions, automates tasks, and connects to your internal systems — without writing a single line of code? That’s exactly what Microsoft Copilot Studio delivers, and it’s more accessible than you might think.

    This guide walks you through what Copilot Studio can do, how to build your first copilot, and where it fits in the broader Microsoft AI ecosystem.

    What is Copilot Studio?

    Copilot Studio (formerly Power Virtual Agents) is Microsoft’s low-code platform for creating custom AI assistants — called copilots. These copilots can:

    • Answer questions using your own data (SharePoint, websites, uploaded files).
    • Follow conversation flows with branching logic, conditions, and variables.
    • Take actions by calling Power Automate flows, APIs, or connectors.
    • Operate across channels: Microsoft Teams, websites, Facebook, Slack, and more.

    The key difference from a simple chatbot: Copilot Studio combines generative AI (GPT-powered answers grounded in your data) with deterministic topics (structured conversation paths you control). You get the flexibility of AI with the reliability of predefined logic.

    Generative answers vs. authored topics

    Copilot Studio gives you two ways to handle user questions:

    ApproachHow it worksBest for
    Generative answersThe copilot searches your data sources and generates a natural language response using AIFAQs, knowledge bases, documentation, general inquiries
    Authored topicsYou design a specific conversation flow with triggers, questions, conditions, and actionsProcesses that require specific steps, data collection, system integrations

    In practice, most copilots use both. Generative answers handle the broad “long tail” of questions, while authored topics manage critical processes where you need full control over the experience.

    What can you build?

    IT help desk assistant

    A copilot that answers common IT questions (“How do I reset my password?”, “How do I connect to the VPN?”) using your internal documentation, and escalates to a human agent when needed.

    HR onboarding companion

    New employees ask about benefits, policies, and procedures. The copilot pulls answers from SharePoint and guides them through onboarding tasks step by step.

    Customer support agent

    Deploy a copilot on your website that handles product questions, checks order status (via API calls), and creates support tickets when it can’t resolve an issue.

    Internal operations bot

    A Teams-based copilot that lets employees submit vacation requests, check remaining PTO, or look up project status — all through natural conversation.

    Building your first copilot

    Step 1: Access Copilot Studio

    1. Go to copilotstudio.microsoft.com.
    2. Sign in with your Microsoft 365 or Power Platform account.
    3. Click Create in the left menu and select New copilot.

    Step 2: Describe your copilot

    Copilot Studio lets you set up your assistant by simply describing what it should do. Provide:

    • A name (e.g., “Contoso IT Helper”).
    • A description of its purpose and behavior.
    • Instructions that guide tone and boundaries (e.g., “Answer only IT-related questions. Be concise and professional. If unsure, suggest contacting the IT help desk.”).

    Step 3: Add knowledge sources

    This is where your copilot gets its intelligence. Add one or more data sources:

    • SharePoint sites: point to your documentation libraries.
    • Public websites: the copilot will crawl and index the content.
    • Uploaded files: PDFs, Word documents, and other files.
    • Dataverse tables: structured business data from your Power Platform environment.

    Once connected, the copilot uses generative AI to answer questions based on these sources — no training required.

    Step 4: Create an authored topic

    For specific processes, create a topic:

    1. Go to the Topics tab and click Add a topic.
    2. Define trigger phrases (e.g., “reset my password”, “I can’t log in”, “password help”).
    3. Build the conversation flow using the visual editor:
      • Ask questions to collect information.
      • Add conditions to branch the conversation.
      • Call a Power Automate flow to take action (e.g., send a password reset link).
      • Display a message confirming the action was completed.

    Step 5: Test and publish

    Use the built-in Test copilot panel to simulate conversations. Once satisfied:

    1. Click Publish in the top menu.
    2. Choose your channel: Microsoft Teams, a website embed (via iframe), or external channels like Slack or Facebook.
    3. For Teams, your copilot appears as a chat app that users can find in the Teams app store.

    Connecting to external systems

    Copilot Studio integrates with 1,000+ connectors through Power Automate. Common integrations include:

    • ServiceNow: create and update support tickets.
    • Salesforce: look up customer information and log activities.
    • SAP: check inventory or order status.
    • Custom APIs: call any REST API using the HTTP connector.
    • Azure OpenAI: run advanced AI prompts beyond the built-in generative capabilities.

    These integrations turn your copilot from a Q&A bot into a genuine virtual assistant that gets things done.

    Copilot Studio vs. Azure Bot Service

    FeatureCopilot StudioAzure Bot Service
    Target audienceBusiness users, citizen developersProfessional developers
    Code requiredNo (low-code visual editor)Yes (C#, JavaScript, Python)
    Built-in AIGenerative answers includedManual integration with AI services
    Connectors1,000+ via Power AutomateCustom code for integrations
    Best forBusiness scenarios, fast deploymentComplex, custom-coded bots

    If you need a copilot up and running fast and your team doesn’t have dedicated developers, Copilot Studio is the right choice. For deeply custom scenarios requiring full code control, Azure Bot Service gives you maximum flexibility.

    Licensing and pricing

    Copilot Studio is licensed per tenant with a capacity-based model:

    • Copilot Studio license: includes 25,000 messages per month per tenant.
    • Additional message packs can be purchased for higher volumes.
    • Microsoft 365 users with certain plans get limited Copilot Studio capabilities as part of their license.
    • A free trial is available at copilotstudio.microsoft.com — no credit card required.

    A “message” is a single interaction (user message + copilot response). Generative AI responses consume more messages than simple authored topic responses.

    Next steps

    1. Start a free trial at copilotstudio.microsoft.com and build your first copilot in under 30 minutes.
    2. Connect a SharePoint site as a knowledge source to see generative answers in action.
    3. Create one authored topic for a specific process your team handles frequently.
    4. Explore the official docs: learn.microsoft.com/microsoft-copilot-studio

    Conclusion

    Copilot Studio puts AI assistant creation in the hands of business teams. You don’t need to be a developer to build a copilot that answers employee questions, automates routine processes, and integrates with the tools your organization already uses. With generative AI handling the broad questions and authored topics managing critical workflows, it strikes the right balance between intelligence and control.

    In the next article, we’ll dive into Semantic Kernel, Microsoft’s open-source SDK for building AI agents and plugins with C# and Python.

    Already thinking about a use case? Tell me what your copilot would do in the comments — I’d love to help you plan it out.

  • Azure AI Document Intelligence

    Every organization deals with paperwork — invoices, receipts, contracts, ID cards, tax forms. Extracting data from these documents manually is slow and error-prone. Azure AI Document Intelligence uses machine learning to read, understand, and extract structured data from documents automatically.

    This article covers what the service offers, which models to use for different scenarios, and how to process your first document with just a few lines of code.

    What is Azure AI Document Intelligence?

    Azure AI Document Intelligence (formerly Form Recognizer) is an AI service that extracts text, key-value pairs, tables, and structures from documents. It handles:

    • Scanned PDFs and images (via built-in OCR).
    • Digital PDFs with complex layouts.
    • Photos of receipts, business cards, and forms.
    • Handwritten text in multiple languages.

    Unlike basic OCR that just reads text, Document Intelligence understands the structure of your documents — it knows which text is a header, which values belong to which fields, and how tables are organized.

    Prebuilt models vs. custom models

    The service comes with two categories of models:

    Prebuilt models

    Ready to use with zero training. Microsoft has already trained these on millions of documents:

    ModelWhat it extracts
    InvoiceVendor name, amounts, line items, tax, due date
    ReceiptMerchant, total, items, date, payment method
    ID DocumentName, date of birth, document number, expiration
    W-2 (US tax)Employer info, wages, tax withholdings
    Health Insurance CardMember ID, group number, plan details
    Business CardName, title, company, phone, email, address
    LayoutText, tables, selection marks, document structure
    ReadPlain text extraction (OCR) with line and word positions

    Custom models

    When your documents don’t match any prebuilt model — like internal forms, proprietary reports, or industry-specific paperwork — you can train a custom model using as few as 5 labeled samples.

    Real-world use cases

    Accounts payable automation

    Process incoming invoices automatically: extract vendor, amount, line items, and PO numbers, then push the data into your ERP system. Teams that used to spend hours on manual data entry can process hundreds of invoices in minutes.

    Expense report processing

    Employees snap photos of receipts. Document Intelligence reads the merchant, date, total, and category, then populates the expense report automatically.

    Contract analysis

    Extract key clauses, dates, and parties from contracts. Combine with Azure OpenAI to summarize terms or flag unusual conditions.

    Healthcare intake

    Read insurance cards and patient forms at check-in. Extract member IDs, group numbers, and patient details to reduce front-desk workload and data entry errors.

    Getting started

    Step 1: Create the resource

    1. Go to the Azure Portal and search for “Document Intelligence”.
    2. Click Create.
    3. Choose your subscription, resource group, region, and pricing tier.
    4. Click Review + Create.

    Step 2: Try it in Document Intelligence Studio

    Before writing any code, explore Document Intelligence Studio at documentintelligence.ai.azure.com. Upload a sample document, pick a prebuilt model, and see the extracted data instantly. It’s the fastest way to evaluate whether a prebuilt model fits your documents.

    Step 3: Analyze a document with Python

    Here’s how to extract data from an invoice using Python:

    from azure.ai.documentintelligence import DocumentIntelligenceClient
from azure.core.credentials import AzureKeyCredential

client = DocumentIntelligenceClient(
    endpoint="https://YOUR-RESOURCE.cognitiveservices.azure.com/",
    credential=AzureKeyCredential("YOUR-API-KEY")
)

# Analyze an invoice from a URL
poller = client.begin_analyze_document(
    "prebuilt-invoice",
    analyze_request={"url_source": "https://example.com/invoice.pdf"}
)
result = poller.result()

for doc in result.documents:
    print(f"Vendor: {doc.fields['VendorName'].content}")
    print(f"Total: {doc.fields['InvoiceTotal'].content}")
    print(f"Date:  {doc.fields['InvoiceDate'].content}")

    if "Items" in doc.fields:
        for item in doc.fields["Items"].value:
            desc = item.value["Description"].content
            amount = item.value["Amount"].content
            print(f"  - {desc}: {amount}")

    Install the SDK:

    pip install azure-ai-documentintelligence

    Step 4: Extract tables and layout

    For documents that don’t match a prebuilt model, the Layout model extracts all text, tables, and structure:

    poller = client.begin_analyze_document(
    "prebuilt-layout",
    analyze_request={"url_source": "https://example.com/report.pdf"}
)
result = poller.result()

# Extract tables
for table in result.tables:
    print(f"Table: {table.row_count} rows x {table.column_count} columns")
    for cell in table.cells:
        print(f"  [{cell.row_index},{cell.column_index}] {cell.content}")

    Combining with Azure OpenAI

    Document Intelligence and Azure OpenAI are a powerful combination. A common pattern:

    1. Extract text and tables from a PDF using Document Intelligence.
    2. Send the extracted content to GPT-4o with a prompt like “Summarize this contract” or “Find all penalties and deadlines.”
    3. Get structured, actionable output that would have taken hours to compile manually.

    This is especially effective for contracts, financial reports, and regulatory filings where you need both extraction accuracy and natural language understanding.

    Pricing overview

    ModelPrice per page (approx.)
    Read (OCR)$0.001
    Layout$0.01
    Prebuilt (Invoice, Receipt, etc.)$0.01
    Custom$0.03 (training is free for first model)

    At these prices, processing 1,000 invoices costs about $10. Compare that to the cost of manual data entry and the ROI becomes obvious.

    Next steps

    1. Open Document Intelligence Studio and upload a real document to see extraction results instantly.
    2. Start with a prebuilt model — invoices and receipts cover the most common automation scenarios.
    3. Train a custom model if your document type isn’t covered — 5 samples is all you need to start.
    4. Read the official docs: learn.microsoft.com/azure/ai-services/document-intelligence

    Conclusion

    Azure AI Document Intelligence eliminates the tedious work of reading and typing data from documents. Whether you’re processing 10 invoices a week or 10,000, the service scales to match your workload. Combined with Azure OpenAI, it turns raw documents into structured, actionable data — the kind of automation that delivers measurable ROI from day one.

    Next up, we’ll explore Copilot Studio, Microsoft’s platform for building custom AI assistants without writing code.

    What documents are slowing your team down? Share your document processing challenge in the comments and let’s figure out the right approach together.

  • Azure AI Search & RAG

    Imagine asking a question in plain English and getting an accurate answer pulled directly from your company’s own documents. That’s the power behind Retrieval-Augmented Generation (RAG), and Azure AI Search is the service that makes it possible at scale on Microsoft Azure.

    In this guide, you’ll learn what Azure AI Search is, how RAG works, and how to build your first search-powered AI solution.

    What is Azure AI Search?

    Azure AI Search (formerly Azure Cognitive Search) is a fully managed search service on Azure. It goes far beyond traditional keyword search by offering:

    • Full-text search: classic keyword matching with filters, facets, and scoring profiles.
    • Vector search: find results based on meaning, not just exact words.
    • Hybrid search: combine keyword and vector search for the best of both worlds.
    • Semantic ranking: a built-in AI layer that re-ranks results by relevance.
    • Integrated vectorization: automatically generate embeddings from your content using Azure OpenAI.

    Think of it as the intelligent retrieval engine sitting between your data and your AI models.

    What is RAG and why does it matter?

    RAG stands for Retrieval-Augmented Generation. It’s a pattern that solves one of the biggest challenges with large language models: they don’t know your private data.

    Here’s how it works:

    1. The user asks a question (e.g., “What’s our refund policy for enterprise clients?”).
    2. The system searches your documents using Azure AI Search to find the most relevant content.
    3. The retrieved content is sent to a language model (like GPT-4o) along with the question.
    4. The model generates a grounded answer based on your actual documents, not its general training data.

    The result? Accurate, up-to-date answers that cite your own sources, with far fewer hallucinations.

    RAG vs. fine-tuning: when to use each

    AspectRAGFine-tuning
    Best forAnswering questions over your documentsChanging the model’s tone, format, or behavior
    Data freshnessAlways current (search index is updated)Frozen at training time
    Setup complexityModerate (index + prompt engineering)High (training pipeline + compute)
    CostSearch service + token usageTraining compute + token usage
    Hallucination riskLower (grounded in retrieved docs)Higher without retrieval

    For most enterprise use cases, RAG is the recommended starting point. Fine-tuning is complementary, not a replacement.

    Key components of a RAG solution on Azure

    1. Data sources

    Azure AI Search can pull data from Azure Blob Storage, Azure SQL Database, Cosmos DB, SharePoint, and many other sources using built-in indexers.

    2. Search index

    Your data is processed and stored in a search index. During indexing, you can apply skillsets that enrich the data — extract text from PDFs, detect languages, split documents into chunks, and generate vector embeddings.

    3. Query pipeline

    When a user asks a question, the query is converted into a vector (using the same embedding model), and Azure AI Search retrieves the most relevant chunks using hybrid search.

    4. Language model

    The retrieved chunks are passed to Azure OpenAI (GPT-4o or similar) as context, and the model generates a natural language answer.

    Setting up your first RAG pipeline

    Step 1: Create an Azure AI Search resource

    1. Go to the Azure Portal (portal.azure.com).
    2. Search for “AI Search” and click Create.
    3. Select your subscription, resource group, and region.
    4. Choose a pricing tier (Free works for testing, Basic for small production workloads).
    5. Click Review + Create.

    Step 2: Upload your documents

    Upload your files (PDFs, Word docs, text files) to an Azure Blob Storage container. This will be your data source.

    Step 3: Create an index with integrated vectorization

    In the Azure portal, use the “Import and vectorize data” wizard on your AI Search resource. It will:

    • Connect to your Blob Storage.
    • Chunk your documents automatically.
    • Generate embeddings using an Azure OpenAI embedding model.
    • Create the search index with both text and vector fields.

    Step 4: Query with Python

    from azure.search.documents import SearchClient
from azure.core.credentials import AzureKeyCredential

client = SearchClient(
    endpoint="https://YOUR-SEARCH-SERVICE.search.windows.net",
    index_name="your-index",
    credential=AzureKeyCredential("YOUR-API-KEY")
)

results = client.search(
    search_text="refund policy for enterprise",
    top=3,
    query_type="semantic",
    semantic_configuration_name="my-semantic-config"
)

for result in results:
    print(result["chunk"], result["@search.score"])

    Step 5: Connect to Azure OpenAI for the full RAG flow

    import openai

# 1. Retrieve relevant chunks (from the search above)
context = "\n\n".join([r["chunk"] for r in results])

# 2. Send to Azure OpenAI with the retrieved context
ai_client = openai.AzureOpenAI(
    api_key="YOUR-OPENAI-KEY",
    api_version="2024-10-21",
    azure_endpoint="https://YOUR-RESOURCE.openai.azure.com/"
)

response = ai_client.chat.completions.create(
    model="gpt-4o",
    messages=[
        {"role": "system", "content": f"Answer using only this context:\n\n{context}"},
        {"role": "user", "content": "What's the refund policy for enterprise clients?"}
    ]
)

print(response.choices[0].message.content)

    To install the required libraries:

    pip install azure-search-documents openai

    Pricing overview

    TierMonthly cost (approx.)Best for
    Free$0Learning and prototyping (50 MB, 3 indexes)
    Basic~$75 USDSmall production workloads (2 GB, 15 indexes)
    Standard S1~$250 USDMedium workloads with semantic ranking

    Vector search and semantic ranking are included at no extra charge on Basic and above. Embedding generation costs depend on your Azure OpenAI pricing.

    Next steps

    1. Try the “Import and vectorize data” wizard in the Azure portal — it’s the fastest way to see RAG in action.
    2. Explore Azure AI Studio: it provides a visual RAG pipeline builder with built-in chat evaluation.
    3. Experiment with chunking strategies: document splitting has a big impact on answer quality.
    4. Check the official docs: learn.microsoft.com/azure/search

    Conclusion

    Azure AI Search combined with RAG lets you build AI solutions that actually know your data. Instead of hoping a language model has the right answer, you give it the right context. It’s the most practical way to bring generative AI into your organization without exposing sensitive data or dealing with hallucinations.

    In the next article, we’ll look at Azure AI Document Intelligence, the service that extracts structured data from forms, invoices, and documents automatically.

    Ready to build your first RAG app? Drop a comment with your use case and I’ll point you in the right direction.

  • GitHub Copilot: Your AI-Powered Coding Assistant

    Imagine having an experienced programmer sitting next to you who suggests code as you type, completes functions for you, and helps you solve problems. That’s exactly what GitHub Copilot does, the most popular AI coding assistant in the world.

    In this article, I’ll explain what it is, how it works, and how you can start using it even if you’re a beginner.

    What is GitHub Copilot?

    GitHub Copilot is an artificial intelligence tool developed by GitHub (owned by Microsoft) that helps you write code faster. It integrates directly into your code editor and offers:

    • Intelligent autocomplete: it doesn’t just complete words, but entire lines and functions.
    • Code generation: describe what you need in a comment and it generates the code.
    • Integrated chat: you can ask questions about your code, have it explain functions, or find bugs.
    • Multi-language support: works with Python, JavaScript, TypeScript, C#, Java, Go, Ruby, and many more.

    How does it work?

    GitHub Copilot uses AI models trained on public code to understand programming patterns. When you write code:

    1. It analyzes the context: it reads the file you’re working on, the comments, variable names, and the project structure.
    2. It predicts what you need: based on that context, it generates code suggestions.
    3. You decide: you can accept the suggestion with Tab, reject it, or ask for an alternative.

    It’s important to understand that Copilot suggests, it doesn’t decide. You always have control and should review what it generates.

    Plans and pricing

    PlanPriceIdeal for
    Free$0/monthStudents and small personal projects (limited suggestions)
    Pro$10/monthIndividual developers
    Business$19/user/monthTeams and companies
    Enterprise$39/user/monthLarge organizations with advanced needs

    The Free plan includes autocomplete and chat with monthly limits. It’s perfect for trying out the tool before committing to a paid plan.

    How to install GitHub Copilot in VS Code

    The most common way to use Copilot is in Visual Studio Code. Here’s a step-by-step guide:

    Step 1: Have a GitHub account

    If you don’t have one, create one for free at github.com.

    Step 2: Activate Copilot on your account

    1. Go to github.com/settings/copilot.
    2. Select the plan you prefer (you can start with Free).
    3. Complete the setup.

    Step 3: Install the extension in VS Code

    1. Open Visual Studio Code.
    2. Go to the Extensions tab (Ctrl+Shift+X).
    3. Search for “GitHub Copilot”.
    4. Click Install.
    5. Also install “GitHub Copilot Chat” for the integrated chat.

    Step 4: Sign in

    1. VS Code will ask you to sign in with your GitHub account.
    2. Authorize the extension.
    3. Done, Copilot is active.

    Copilot in action: practical examples

    Example 1: Intelligent autocomplete

    You type the start of a function:

    def calculate_tax(subtotal, rate):

    Copilot automatically suggests:

    def calculate_tax(subtotal, rate):
    tax = subtotal * (rate / 100)
    return round(tax, 2)

    You press Tab to accept and keep working.

    Example 2: Generate code from a comment

    You write a comment describing what you need:

    # Function that reads a CSV file and returns a dictionary
# where the keys are the column names

    Copilot generates:

    import csv

def read_csv_as_dictionary(file_path):
    with open(file_path, 'r', encoding='utf-8') as file:
        reader = csv.DictReader(file)
        return [row for row in reader]

    Example 3: Use the chat to understand code

    You select a block of code you don’t understand, open the Copilot chat (Ctrl+Shift+I), and ask:

    “What does this code do?”

    Copilot gives you a clear explanation in plain language, line by line if needed.

    Example 4: Ask it to fix a bug

    If you have a bug in your code, you can select it and tell the chat:

    “This code throws a TypeError, can you fix it?”

    Copilot analyzes the problem and suggests the fix.

    Copilot Chat: your pair programming buddy

    Beyond autocomplete, GitHub Copilot includes a chat that you can use to:

    • Explain code: “Explain this function to me”
    • Generate tests: “Generate unit tests for this class”
    • Refactor: “Simplify this function”
    • Document: “Add docstrings to these functions”
    • Debug: “Why does this code return None?”
    • Learn: “What’s the difference between a list and a tuple in Python?”

    To open the chat in VS Code:

    • Side panel: click the Copilot icon in the sidebar.
    • Inline chat: press Ctrl+I inside the editor to request changes directly in your code.

    Tips to get the most out of it

    1. Write descriptive comments

    Copilot works better when it understands your intent. A clear comment like # Validate that the email has the correct format generates better code than just starting to write the function.

    2. Give it context with good names

    Use descriptive names for variables and functions. calculate_total_with_discount(price, percentage) gives Copilot much more context than func(a, b).

    3. Always review the generated code

    Copilot is impressive, but it’s not infallible. It can generate code with subtle bugs, security vulnerabilities, or that simply doesn’t do what you expect. Always review before accepting.

    4. Use the chat to learn

    If you’re a beginner, Copilot’s chat is an incredible learning tool. It doesn’t just give you code, it explains why it works that way.

    5. Try the Ctrl+Enter shortcut

    If the first suggestion doesn’t convince you, press Ctrl+Enter to see multiple alternatives and choose the one that fits best.

    Other compatible editors

    While VS Code is the most popular, GitHub Copilot also works in:

    • Visual Studio (Microsoft’s full IDE)
    • JetBrains IDEs (IntelliJ, PyCharm, WebStorm, Rider, etc.)
    • Neovim
    • Xcode (beta support)

    Limitations you should know about

    • It doesn’t understand your full business logic: Copilot sees individual files and nearby context, not your entire architecture.
    • It can suggest insecure code: don’t blindly trust suggestions for code that handles authentication, encryption, or sensitive data.
    • It’s not always correct: suggestions are probabilistic, not deterministic. The same prompt can give different results.
    • Licensed code: although GitHub has taken measures, there’s an ongoing debate about whether some suggestions may reflect code with specific licenses.

    Conclusion

    GitHub Copilot has changed the way millions of developers work. Whether you’re just starting to code or you’ve been at it for years, having an AI assistant that understands your code and helps you in real time is an advantage that’s hard to ignore.

    My recommendation: activate the free plan, install the extension in VS Code, and use it for a week. You’ll notice the difference from day one.

    Do you use GitHub Copilot? What’s your favorite trick? Share it in the comments.

  • Microsoft Copilot: A Complete Guide for Beginners

    Microsoft has gone all in on artificial intelligence, and its most visible move is Copilot: an AI assistant integrated into virtually every Microsoft product. But with so many versions and names, it’s easy to get lost. In this article, I’ll explain what Copilot is, what its variants are, and how you can start using it today.

    What is Microsoft Copilot?

    Microsoft Copilot is an artificial intelligence assistant powered by large language models (like GPT-4) that is integrated into Microsoft products. Its goal is to help you be more productive: drafting text, analyzing data, creating presentations, automating tasks, and much more, using natural language.

    Instead of learning complicated commands, you simply tell Copilot what you need and it does it for you.

    The different versions of Copilot

    This is where many people get confused. Microsoft uses the name “Copilot” for several different products:

    Copilot (free)

    • Available at copilot.microsoft.com and in the Windows 11 sidebar.
    • Works like Microsoft’s version of ChatGPT: you can ask questions, generate text, create images, and have conversations.
    • No paid subscription required.

    Microsoft 365 Copilot

    • Integrated directly into Word, Excel, PowerPoint, Outlook, and Teams.
    • Requires a Microsoft 365 license and an additional Copilot subscription.
    • The most powerful option for work environments because it has access to your documents, emails, and calendar.

    Copilot Studio

    • A platform for creating custom AI agents without needing to code.
    • Allows you to connect your own data sources and define conversation flows.
    • Ideal for companies that want specialized chatbots.

    GitHub Copilot

    • Designed specifically for developers.
    • Integrates into code editors like VS Code.
    • We’ll cover it in detail in the next article.

    What can Microsoft 365 Copilot do?

    Let’s look at concrete examples in each application:

    In Word

    • “Write a 500-word report on AI trends in 2025” and Copilot generates the complete draft.
    • “Summarize this document in 3 key points” and you get an instant summary.
    • “Change the tone of this paragraph to more formal” and it rewrites the text.

    In Excel

    • “Create a pivot table with sales by region” and Copilot generates it automatically.
    • “Which month had the highest revenue?” and it analyzes the data for you.
    • “Add a column with the monthly growth percentage” and it writes the formula.

    In PowerPoint

    • “Create a 10-slide presentation about our strategic plan” using content from a Word document.
    • “Add a slide with a comparison chart” and it generates one.
    • “Improve the design of this presentation” and it applies visual changes.

    In Outlook

    • “Summarize the emails from this week about Project Alpha” and it gives you a summary.
    • “Draft a professional response declining this meeting” and it prepares the email.
    • “What pending items do I have based on my recent emails?” and it extracts the tasks.

    In Teams

    • “Summarize what was discussed in today’s meeting” (from the transcript).
    • “What decisions were made?” and it extracts the key points.
    • “Draft a follow-up for the team” with the action items.

    How to start using Copilot (free)

    The fastest way to try Copilot without spending anything:

    Option 1: In the browser

    1. Go to copilot.microsoft.com.
    2. Sign in with your Microsoft account (or use it without one).
    3. Type your question or request.

    Option 2: In Windows 11

    1. Press Windows + C (or search for “Copilot” in the Start menu).
    2. The Copilot side panel opens.
    3. You can ask it to change system settings, find files, or answer questions.

    Option 3: On the mobile app

    1. Download the Microsoft Copilot app on iOS or Android.
    2. Sign in and start using it as a personal assistant.

    How to access Microsoft 365 Copilot

    For the version integrated into Office, you need:

    1. A Microsoft 365 license (Business Standard, Business Premium, E3, or E5).
    2. The Copilot subscription: $30 USD/user/month for businesses.
    3. Activation by your organization’s administrator.

    If you’re a small business or freelancer, evaluate whether the cost is justified by the time you’ll save. For many users, Copilot in Word and Excel alone justifies the investment.

    Tips to get the most out of it

    1. Be specific in your instructions

    Instead of: “Make me a presentation”
    Better: “Create an 8-slide presentation about the benefits of migrating to the cloud, targeted at IT directors, with data and charts”

    2. Provide context

    Copilot works better when it has context. In Word, select a paragraph before asking it to improve it. In Excel, make sure your data has clear headers.

    3. Iterate

    The first result won’t always be perfect. Ask for adjustments: “Make it shorter”, “Add examples”, “Shift the focus to costs”.

    4. Combine tools

    The real power of Copilot shows when you use it across applications. For example:

    • Generate an analysis in Excel.
    • Ask Copilot in Word to create a report based on that analysis.
    • Then generate a presentation in PowerPoint with the key takeaways.

    Limitations you should know about

    • It can generate incorrect information: always verify important data, especially numbers and dates.
    • It doesn’t replace human judgment: it’s an assistant, not a decision-maker.
    • It works better in English: while it supports other languages, English responses tend to be more accurate.
    • It requires well-organized data: in Excel, if your data is messy, Copilot’s results will be too.

    Conclusion

    Microsoft Copilot is one of the most accessible AI tools on the market. From the free browser version to the full Microsoft 365 integration, there’s an option for every type of user. The key is to start using it, experiment with different prompts, and discover how it can save you time in your daily work.

    In the next article, we’ll focus on GitHub Copilot, the AI tool that is transforming the way developers write code.

    Have you tried Copilot? Share your experience in the comments.

  • What is Azure OpenAI Service and How to Get Started

    If you’ve heard about ChatGPT, you’re probably wondering how companies are using that same technology in their own projects. The answer is Azure OpenAI Service, Microsoft’s cloud service that lets you use OpenAI models (like GPT-4, GPT-4o, and DALL-E) within Azure, with all the security and compliance that organizations need.

    In this article, I’ll explain what it is, what it’s used for, and how you can take your first steps.

    What is Azure OpenAI Service?

    Azure OpenAI Service is a cloud service on Microsoft Azure that gives you access to OpenAI’s artificial intelligence models. This includes:

    • GPT-4o and GPT-4: for generating text, answering questions, summarizing documents, and much more.
    • DALL-E 3: for generating images from text.
    • Whisper: for converting audio to text.
    • Embeddings: for semantic search and recommendation systems.

    The key difference from using OpenAI directly is that Azure OpenAI offers:

    • Enterprise security: your data is not used to train models.
    • Regulatory compliance: certifications like ISO 27001, SOC 2, HIPAA, and more.
    • Private networking: you can connect the service to your Azure virtual network.
    • Access control: integration with Azure Active Directory (Entra ID).

    How is it used in the real world?

    Intelligent chatbots

    Build virtual assistants that understand natural language and can answer questions about products, services, or internal documentation.

    Content generation

    Automate the writing of emails, reports, product descriptions, or blog posts.

    Document analysis

    Summarize contracts, extract key information from reports, or automatically classify support tickets.

    Smart search

    Implement semantic search over internal knowledge bases, where users ask questions in their own words and the system finds the relevant answer.

    What do you need to get started?

    1. An Azure subscription

    If you don’t have one, you can create a free account at azure.microsoft.com. Microsoft gives you $200 USD in credits for 30 days.

    2. Request access to Azure OpenAI

    Azure OpenAI requires prior approval. You need to fill out a form in the Azure portal indicating your use case. Approval usually takes between 1 and 5 business days.

    3. Create the resource in Azure

    Once approved:

    1. Go to the Azure Portal (portal.azure.com).
    2. Search for “Azure OpenAI” in the search bar.
    3. Click Create.
    4. Select your subscription, resource group, and region.
    5. Give the resource a name and select the pricing tier.
    6. Click Review + Create.

    4. Deploy a model

    Inside your Azure OpenAI resource:

    1. Go to Azure OpenAI Studio (oai.azure.com).
    2. In the side menu, select Deployments.
    3. Click Create new deployment.
    4. Choose the model (for example, gpt-4o).
    5. Name the deployment and confirm.

    Your first API call

    Once you have your model deployed, you can test it directly from Azure OpenAI Studio using the Chat Playground. But if you want to do it from code, here’s a Python example:

    import openai

client = openai.AzureOpenAI(
    api_key="YOUR_API_KEY",
    api_version="2024-10-21",
    azure_endpoint="https://YOUR-RESOURCE.openai.azure.com/"
)

response = client.chat.completions.create(
    model="your-deployment-name",
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "What is artificial intelligence?"}
    ]
)

print(response.choices[0].message.content)

    To install the required library:

    pip install openai

    How much does it cost?

    Azure OpenAI charges per token (chunks of text). As an approximate reference:

    ModelInput (per 1M tokens)Output (per 1M tokens)
    GPT-4o~$2.50 USD~$10.00 USD
    GPT-4o mini~$0.15 USD~$0.60 USD
    GPT-4~$30.00 USD~$60.00 USD

    For a small or test project, costs are usually just a few dollars per month. You can set up spending alerts in Azure to avoid surprises.

    Next steps

    Now that you know the basics of Azure OpenAI Service, I recommend:

    1. Create your free Azure account and request access.
    2. Explore Azure OpenAI Studio: the playground lets you experiment without writing code.
    3. Try a simple use case: a basic chatbot or a summary generator.
    4. Check out the official documentation: learn.microsoft.com/azure/ai-services/openai

    Conclusion

    Azure OpenAI Service is the gateway to bringing generative AI into your projects and your business, with the trust and security that Microsoft Azure provides. You don’t need to be an AI expert to get started: with an Azure account and a few minutes of setup, you can have your first model up and running.

    In the next article, we’ll explore Microsoft Copilot, the AI assistant that Microsoft is integrating into all of its tools.

    Have questions? Leave a comment and I’ll help you take your first steps with Azure OpenAI.

  • Hello world!

    Welcome to WordPress. This is your first post. Edit or delete it, then start writing!